Ethambutol (EMB) was developed at Lederle Laboratories in the late 1 950s from a collection of about 2000 mostly symmetrical diamines. Ethambutol is an ideal small-molecule drug to be optimized by combinatorial chemistry: it has low toxicity and good pharmacokinetics, but a relatively high MIG of 5 mg/ml that could be dramatically improved through additional analoging and medicinal chemistry. We have devised a protocol for solid-phase synthesis of asymmetric 1 ,2- diamines related to ethambutol and used this protocol to synthesize a test library of 100,000 structural analogs. This library has been screened using both a whole-cell luciferase system that responds positively to cell- wall inhibition and using microbroth dilution assays for the determination of MIGs against Mycobacterium tuberculosis. From this procedure we arrived at approximately 300 discrete hits that show activity in both assays. We propose to extend the structural diversity represented-by this library with particular emphasis on classes of molecules underrepresented in the first library as well as additional analoging around viable hits. Medicinal chemistry will be guided by both SAR and likelihood of intestinal absorption. We propose a series of biochemical analyses that will allow us to prioritize hit molecules and confirm that these molecules share a common target in order to facilitate interpretation of SAR. These assays will be extended to include preliminary pharmacologic and pharmacokinetic analysis of related lead series. We will further establish in vivo screening systems that will allow us to move appropriate hit series forward into small animal models of tuberculosis. These studies should allow us to identify candidates that will ultimately be evaluated for efficacy in humans.